Over the last few months, I’ve heard from a dozen or more community professionals who report a significant decline in the number of visitors to their community.
This raises a question: Are fewer visitors reaching posts in our communities, or has the level of engagement also declined?
To address this, we’ve scraped data from 27 communities, analyzed the number of replies to posts, and segmented the communities into three buckets.
- High-Volume: More than 1k+ replies per month (12 communities).
- Mid-Volume: 100 to 1k replies per month (10 communities).
- Low-Volume: Less than 100 replies per month (5 communities).
This is hardly a scientific measure, but it might serve as a helpful indicator of where things are heading. These are the combined results:
In all three tiers, the level of participation has declined somewhat over the past four years.
However, it’s the engagement in more recent months that raises a significant concern.
The year-on-year data, particularly for the past few months since March (when Google rolled out the new core update), has shown a significant decline in engagement of between 20% and 40%.
And this appears to be impacting almost all communities equally.
The Three Headwinds We’re Facing
My best guess is this represents three headwinds colliding into something close to a perfect storm.
- A slow rise in using LLMs for questions that were previously asked in a community. Even today, I find myself using ChatGPT for answers vs. asking in a forum. The immediacy is a win.
- An ongoing shift away from forum-centric questions to other platforms. There is a slow but steady drift away from hosted discussion-based platforms to different places to ask questions and get expert advice.
- Google showing community answers in AI overviews. This means fewer people need to visit a community to get answers.
The Google Search Update Change Is Causing Participation To Drop
Of the three, Google’s Search updates seem to be having the most significant impact.
In the past, you would search for an answer, find a title that seemed feasible in the search results, and then visit it. If that solves your question, great; if not, you can ask a question in the community (since you’re already there).
Now Google is displaying answers in search, people don’t land on a community and aren’t prompted to ask the answer there (it’s not quite clear where they’re going instead – my best guess is they browse a few more search results and either file a ticket, ask in Reddit, or simply give up).
The trend here is quite clear – almost all platforms are keeping more traffic for themselves.
And if you follow this trend to its end destination, it’s worthwhile considering the most extreme example: Google Zero – I.e., you get zero traffic from search engines.
This is the scenario that we need to begin preparing for (even if it doesn’t happen, it’s worth preparing for a world of ever-declining traffic volume).
Some Communities Are In Trouble, Others Not So Much
The first challenge is determining if your community is in trouble or not.
The answer to this question depends on two things.
- How reliant is your community on search traffic for discovery? Those who rely on search traffic for visitors are obviously most susceptible to changes in search engine algorithms and how they pull and display information.
- How much competition is there for the value your community provides? Can other channels realistically give the same value that the community offers? If you just want information, numerous channels are available for that purpose. But what if you want something more?
The answers to these questions will determine how susceptible you are to changes
Transactional Communities Might Face A Bumpy Few Years
As I’ve written before, most enterprise communities are transactional.
And transactional communities, which are heavily reliant on search for discovery, will also be most susceptible to sudden changes when Google tweaks its algorithm.
So, are transactional communities doomed?
No, but they probably won’t reach the levels of participation in the past (at least not mature communities).
However, transactional communities will survive because their customers will always have questions for which search engines and LLMs can’t provide a satisfactory answer.
This means that even if Google Zero occurs (or search traffic continues to decline), people still have questions that need answers. Customer support representatives can only support a limited number of people (and LLMs can only provide answers that already exist in a community). This means that there will also be a need for people to ask questions and receive answers from their peers.
This means there’s a baseline number of questions that people will always need answers to.
The Google Zero Experiment
Coincidentally, this thought experience (zero traffic from Google) is something we’ve tried before (though to prove ROI vs. measuring participation declines).
When the community was removed from search engine results, the number of support tickets increased significantly, while there was still a baseline number of visitors and questions being published in the community.
Clearly, the total number of questions was much lower. Most people simply filed a support ticket.
However, what if we did more work to deflect people about to file a support ticket to the community? It wouldn’t take much to deflect those about to file a support ticket to the community first.
And even then, by ramping up awareness and promotional efforts to customers, especially via CSMs, it’s not hard to imagine the community still having a solid and healthy level of engagement.
But the key thing is this:
Many purely transactional, support-centric communities will likely never regain their former level of activity. This will have an impact on everything from superuser programs, gamification, user groups, etc, but they will still survive and deliver the same value as before.
(Aside: every platform vendor charging by visit instead of some measure of participation is in for an especially rough couple of years)
At the other extreme are identity-based communities. These are the communities that bring people together via a desire to build relationships with one another. Most peer groups fall into this category.
This group is unlikely to be impacted because they don’t rely on search traffic, and there is limited competition for the unique relational value they provide.
(Aside – Most private communities and paid membership communities should be fine for the same reason.)
Most peer groups and user groups fall squarely into this category. People will always want to build relationships with one another, and there aren’t many places that offer this opportunity. It’s not something that LLMs and other trends will impact.
I suspect this is where the notion-style approach becomes most powerful.
The two community categories which are in trouble are those focused on a common interest and those around specific goals.
These are simultaneously impacted by both the change in search engine algorithms and how people gather information. These are the broad communities of interest and goal-specific communities of practice. They have neither the protection of the baseline number of questions to resolve nor offer unique value that people might use LLMs or other channels for.
It’s these communities that are most susceptible to a death spiral – as engagement declines, fewer people participate with increasingly less frequency. Without the baseline and/or search traffic, people simply go elsewhere to satisfy their needs.
And I suspect it’s these communities that are going to really struggle.
What Should You Do?
The easiest thing to do is to go where members go. This is the Community Everywhere approach. You simply follow members on their journey. You map out the existing landscape, identify what already exists, discover how to add value, and follow the new playbook for building communities.
The problem is that this isn’t always easy to do.
Many of us have invested too much time and energy into hosted communities. And most organisations simply don’t have the structure or ability to embrace a Community Everywhere approach.
This means that many of us are responsible for hosting a community in an environment that is increasingly challenging. It’s not an impossible challenge, but it does mean adapting in key areas.
Transactional Communities > Change The Flow And Measurement
Mature transactional communities should prepare for three things:
- Expect a decline in engagement until we hit a core baseline number of questions. Again, this doesn’t mean your community is delivering less value; you just need to change how you measure it. You need to measure ‘ghost deflections’ instead of just those who get an answer from your website.
Prepare for ‘Google Zero’ – a world where you get zero traffic from search engines. In which case, you need to invest a lot of time in redirecting those who would otherwise contact support to contact you. The Okta support experience (a former client) serves as a good example, but even this could be clearer about which type of questions belong where.
Notice above how a subtle tweak to the image above can have a significant impact on which channels people use.
Finally, ensure that your share of the budget aligns with your share of the discussions. Track what % of questions are being asked on other platforms (especially Reddit), and decide whether to engage there too. You’re not abandoning your hosted platform, but you’re not ignoring shift trends either.
Interest-Based Communities > Become Destination Sites
For those running interest-based communities, you need to become destinations.
This means you need a ‘pull factor’ far beyond ‘there’s conversations about this topic happening here’.
This accounts for the majority of success-centric communities, but could also apply to many communities of practice. This means a heavy focus on things that only your community is delivering.
The best approach here is events.
Sure, hosting events is hardly a new idea. The problem is that most organisations pick the safest and dullest possible events and host them repetitively (guest speaker/interviews). They double down on precisely the most tedious type of events.
Yet, there is a wide range of possibilities to host a variety of events.
As you can see here, the options include the typical knowledge-sharing events. But also events which are far more exciting and engaging to host. Challenges, influencer-partnerships, live debates, show and tell, ‘what I did wrong?’ etc, etc…
Go wild with creativity here. Brainstorm (or use LLMs) to give you a list of the most whackiest, exciting, unique events you can host – and host them. Test and continually refine what works and what doesn’t.
The key isn’t just to host events, but to host events that are so exciting that they attract attention and spread by word of mouth.
Discussions might decline, but the community would remain a lively and engaging place. Only the focus has shifted from discussion-centric to event-centric.
Goal-Based Communities > Identity Communities
The final approach is to prioritize relationship-building activities over discussions.
The key here is to return to our roots and build a genuine sense of community among smaller groups of members. This involves focusing on user groups and peer groups, while also fostering a strong sense of community.
But it’s hard (not impossible, mind) to build a sense of community amongst a large group. Which means the shift will be building local chapters or industry-specific sub-groups, as well as cohort-led groups, to foster a strong sense of community among them.
This could be hosted on a central platform, which can work well, or it might follow the Community Everywhere approach, similar to Notion below.
Notice that Notion isn’t running all these communities; it’s simply providing both the links to groups (on any platform) and solving the discoverability problem.
The Future For Hosted Communities
We can boil this down into four key points.
- Follow Community Everywhere if you can. The easiest way to survive the tailwinds is to let them carry you on their path. This means follow the new playbook, map out the ecosystem and add value to it.
- Change how you measure and attract visitors. This requires anticipating ‘Google Zero’ and redirecting from the support ticket flow, an emphasis on resolving questions both within and outside the community, and a change in how we measure community success.
- Event-centric interest communities. This means transforming communities into destinations with unique, fun, and valuable events that attendees will enjoy. The more distinct, the better.
- A shift to user groups and chapter models. Creating places where members can feel a high sense of community and belonging to one another (vs. the vast open expanses many encounter today).
The glory days might be over, but the future is still bright, and communities remain as indispensable in the future as they’ve ever been.
